What Emergency Medicine Can Teach You About Software Development
summary
Shows concrete analogies: thermography proves a typical emergency blanket fails to stop conductive heat loss; ETCO2 proves tube placement and beats guesswork; experiments find 50% missed the gorilla in a visibility test and 83% of radiologists missed the same gorilla in CTs, warning against tunnel vision. Calls out costly choices — Spring DI, ORMs, microservices — and operational fixes: dry-run, read-only DBs, colored terminals, checklists (ABC/SAMPLE), point-and-call, and postmortems after a two‑hour failed rollback.
AI-written summary · about
chapters (21)
- 0:00Introduction and speaker background
- 1:06Why emergency medicine matters
- 3:11Cargo cult in software development
- 6:00Framing the problem with frameworks
- 10:37Proving solutions with metrics
- 14:47Tunnel vision in debugging
- 18:04Convenient diagnosis in code
- 19:07Golden hammer anti-pattern
- 21:51System design flaws
- 23:14Security layers and Swiss cheese model
- 25:20Read-only access best practices
- 26:42Environment distinction techniques
- 27:46Dry run commands for safety
- 28:47Importance of clear naming
- 31:13Using checklists for reliability
- 36:23Handling errors and retrospectives
- 39:13Planning for failures
- 41:40Writing effective post-mortems
- 43:27Building a safety culture
- 46:55First aid takeaways and conclusion
- 47:58Q&A about debugging approaches
related talks
29mLightning Talks
15mWhen Systems Flatline—Enhancing Incident Response with Learnings from the Medical Field
14mThe Legend of the Arsonist Firefighter
30mBehavioral Software Engineering
25mMental Models for SREs
28mOne weird trick to becoming a better software developer
54mSeeing Like a Programmer - Resiliency, Limits, And Moral Hazards In SE
34mDevEmoOps: An emotionally integrated approach to the software development lifecycle